Pakistan’s artificial intelligence policy lacks – intelligence

    Minute Mirror - Subscribe
    Minute Mirror - Subscribe

    Recently, the Ministry of IT & Telecom (MoITT) released Pakistan’s first AI policy for public consultation. While the document represents an evolution from previous IT policies, it lacks the brevity and boldness required to address the country’s challenges effectively. The draft fails to establish clear accountability and responsibility matrices, particularly concerning the allocation of work and leadership between academia and the government. It sets ambitious goals of training a massive workforce of ten thousand trainers to educate one million AI graduates by 2026. However, it overlooks the need for a robust commercial and technological AI infrastructure to foster job creation and attract investments.

    In an article published in The Friday Times, one of the significant concerns is the absence of legal safeguards for commercial and personal information, hindering potential investments. The draft introduces bureaucratic hurdles through the Centers of Excellence on AI (COE-AI) and an AI regulatory directorate, funded by a risky ‘AI Fund.’ Unfortunately, it fails to prioritize national sectors and lacks measurable targets. For instance, finding fifty IT-enabled municipalities may prove more challenging than implementing AI in those areas by 2030. Similarly, the creation of a smart citizen portal for municipal services falls under the jurisdiction of provincial IT boards, rendering the establishment of COEs redundant.

    Regarding statutory infrastructures, the draft proposes a data protection statute and the formation of a National Commission on Personal Data Protection (NCPDP) by 2024. However, the political climate makes it highly unlikely for the data protection bill to pass, leaving commercial and personal information vulnerable and impeding real investments.

    The document also overlooks the role of different regulators in fostering innovation and establishing a conducive regulatory environment. Currently, applications collecting data from local subscribers operate without domestic legal oversight, while local data centers and cloud infrastructure lack proper regulation. It is essential to involve regulators such as PTA, SBP, PEMRA, and NADRA to set realistic priorities and objectives.

    While the draft rightly emphasizes data standardization in the public sector, the current political environment makes it a sensitive issue. Additionally, the government’s target of filing two thousand patents and registering one hundred commercial models annually by 2026 seems unrealistic considering Pakistan’s track record.

    However, there are opportunities for successful incubators like the National Incubation Center and various academic facilities. These incubators should be empowered with additional computational resources and become centers of excellence that compete for funding based on achievements rather than grants. They can also serve as sandboxes for partnerships with international players like Google and Microsoft. PTA can act as the secretariat and coordinator for these activities, transforming into a technology regulator beyond its current scope.

    To enable these changes, there must be a shift in the approach to technology appointments, moving away from political considerations and favoritism. IT infrastructure should be regulated with the same credibility and firmness as the financial sector. As data becomes increasingly valuable, capacity building in the regulatory space is crucial. The government should relinquish control and allow regulators and industry experts to define the path forward.

    In light of the numerous planning flaws and verbosity in the current policy draft, Sun Tzu’s words from The Art of War come to mind: “If there is no skill in planning, it is difficult to achieve it, and if there is no skill in planning, it will fail.”